Originally published on July 17, 2021
The first half of 2021 picked up right where the last half of 2020 ended: stellar returns in stocks and strong economic numbers. While risks are building in the stock market, momentum is a stubborn bull to fight. And right now, momentum continues to push stock markets to new highs.
Figure 1 shows annualized returns for various indices over various time periods. It is always good to have a healthy dose of perspective, and this chart gives us just that.
Figure 1: Annualized Returns for Various Asset Classes
It is easy to cherry-pick a time horizon as a basis for arguing returns have been strong (or weak). Time horizons have a drastic impact. For example, let’s focus on the S&P 500. YTD, the return of 15.25% is phenomenal; it is well above average. Ditto for the 5-year and 10-year averages. But what happens when you go 20 years out: the average falls to 8.60%. One can justifiably argue that the index is overvalued because its 5- and 10-year returns have been so strong. But one can also justifiably argue that the index is fairly valued (maybe even undervalued), based on the much-less-impressive 20-year return. After all, using data from slickcharts.com (https://www.slickcharts.com/sp500/returns), the average S&P 500 total return since 1926 (earliest available date for data) has been 10.41%.
Included in this chart are returns since top of the market prior to the Great Recession (10/09/2007) and the bottom of the market during the Great Recession (03/09/2009). It is no surprise that returns since the bottom have been quite strong. It is also no surprise that returns since the top of the market before the recession are considerably weaker. Which is a better point from which to measure? It is relatively arbitrary, as we could make the case for either. Again, time horizons make a big difference. Perspective matters.
So how do we know if stocks (and bonds) are overvalued or undervalued if it appears, based on time horizons alone, that the argument could go either way? Feast on the charts below as we dive into this. But first, let us talk about bonds.
Bonds
It is unusual – very unusual – for bonds to have negative returns. In fact, since 1980, there have only been three years when bonds had negative total returns.
As Figure 2 highlights, $100,000 invested into the Barclays US Aggregate Bond index in January 1980 would have grown to $1,970,000 by June 30, 2021.
Steady as that growth has been, it was not without blemish. Circled in red are the years when the index was negative. These years were 1994, 1999, 2013, and the first half of 2021. In 40+ years of history, only three negative years. Not bad. Especially not bad considering that those negative years were relatively mundane compared to the negative years in the stock market.
Figure 3 shows the returns for the three negative years as well as returns for the following years. While the sample size is small, bonds have been positive each year after a negative year, and those positive next-year returns tend to be larger than the average annualized return (e.g., 7.45%) of bonds since January 1980.
To support the notion that future bond returns tend to be strong after a negative period, Figure 4 supplies a wealth of great data. Let’s break it down:
Upper left chart (“US 30-Year Bond Yield”) – This chart shows the interest rate on the US 30-year Treasury Bond over the last three decades.
Lower left chart (“12 Month Actual Change in 30-Year Bond Yields”) – This chart shows the annual change in the interest rate for the 30-year Treasury. For example, the interest rate went from around 6% in 1994 to around 8% in 1995. This was a +2% change year-over-year. This 2% change is reflected on this chart for 1995. NOTE: There are seven periods during which the 30-year rate changed more than +1% year-over-year
Upper right chart (“Fwd Change in 30Y Bonds Yields w/ YoY Cross>1% Last 7X ’90-’13) – For each of the seven periods during which rates increased by at least +1%, this chart shows how the rate changed moving forward after each of those seven periods. For example, after the +2% change from 1994 to 1995, rates fell -0.20% over the next month, -0.18% over the next three months, etc. NOTE: For all seven scenarios, future rate changes – from one month into the future through two years into the future, were overwhelmingly negative.
Recall that bond prices go up (↑) when interest rates fall (↓). Conversely, bond prices go down (↓) when interest rates rise (↑). Thus, bond investors cheer for falling rates.
To this point, notice the three negative years (↓) from Figure 3 were all years where 30-year rates increased (↑) by at least +1% from the previous year from Figure 4. After this happened, 30-year rates tended to fall moving forward (↓), and bond returns were positive (↑).
We counsel patience with bonds now. As we will discuss later, future bonds returns are expected to be higher than future stock returns. As such, an allocation to bonds now may prove shrewd moving forward.
A final comment on bonds. There is much talk about rising interest rates and rising inflation. Of course, we know that rising interest rates are not favorable to bond prices. And similarly, rising inflation is not friendly to bonds. Given concerns about rising rates and rising inflation, how might bonds react long-term?
Figure 5 and Figure 6 below show the same data. Figure 5 shows consumer price index (e.g., inflation) on the left axis and the US Treasury 10-Year Bond interest rate (e.g., interest rates) on the right axis. Figure 6 shows the exact same data, except it zooms in on the period from 1970 through 1983.
In Figure 5, we added the green line to highlight how quickly inflation accelerated during the mid-1970s. The slope (e.g., increase) of inflation during this period is drastic compared to the period before and after it. To boot, during this same period, the 10-Year interest rate climbed from around 7.5% to over 15%! Rates doubled in a little over one decade. Talk about a double-whammy sucker punch to bonds.
Bonds would be expected to perform relatively poorly during this time. The data, however, paints a decidedly different picture. We dug through the Morningstar database to find high quality bond mutual funds that existed during this period. We then analyzed the returns of these bond mutual funds during this period. Figure 7 has the results.
Recall that the total return earned on any investment is the sum of two parts:
Total Return = Principal Increase/Decrease + Income (e.g., Dividends & Interest)
During a period of acute rising rates and rising interest, bonds funds had positive total returns. Assuming $100,000 invested in each fund at the beginning of the period, each fund returned at least 5.09% per year through December 1981. The fascinating part is how the returns were earned:
The yellow column represents the principal increase/decrease portion of total return. As expected, rising rates and inflation (↑) put downward pressure on bond prices (↓). In all cases, the initial principal declined in value. BUT…
The orange column represents the income portion of total return. Despite pressures to bond prices, those bonds paid interest over the entire period. As that interest was reinvested, it grew drastically within the fund.
The green column represents total return, or the sum of the yellow and orange columns.
So, while it is true the rising rate and inflation put pressure on bond prices, it is also true the bond income can offset those effects.
Armed with the knowledge that it is highly unusual for bonds to have negative annual returns; that periods of rapid rate increases (price decreases) tend to be followed by periods of rate decreases (price increases); and that bonds can generate positive total returns even in periods of increasing rate and inflation, we have constructed an argument for bonds to continue providing long-term positive total returns moving forward.
Now the question is: How will bonds fare relative to stocks?
Equities
For some insight into stocks relative to bonds, we turn to the equity risk premium model developed by Hussman Strategic Advisors. For context in this analysis, risk premium is defined as:
Risk Premium = Stock Return – Bond Return
Thus, risk premium is positive if stock returns > bonds returns, and risk premium is negative if stock returns < bond returns.
Figure 8 shows the risk premium model. A few important points:
The blue line shows the predicted risk premium (stock - bonds) over the next 12 years. For example, as of June 2021, the risk premium is -7.0%. Thus, it is predicted that bonds will outperform stocks by 7.0% per year over the next 12 years (remember, a negative premium means bonds > stocks).
The red line shows the actual risk premium over time. Whereas blue is a prediction, red is actual. Red measures the risk premium over the next 12 years. For example, as of 2009, the actual risk premium was 12% (stocks > bonds). This means that from 2009 to 2021, stocks overperformed bonds by 12% per year. Note: Since the 12-year period ending 2022 has not yet been completed, there is no red line for 2010; since the 12-year period ending 2023 has not yet been completed, there is no red line for 2011,…
For the blue prediction line to have any credibility, the red actual line needs to closely trace the blue prediction line. And it does. This model goes back to 1928, and since that time, actual risk premium moves in lockstep with predicted risk premium. In other words, the prediction has credibility.
Long story short, if this model continues to have good predictive ability, bonds may outperform stocks over the next 12 years.
That result may be surprising. Earlier in this piece, the question was asked if stocks are overvalued or undervalued. We believe this surprising risk premium prediction is largely driven by the historical overvaluation in the stock market today. The following charts shed some perspective on this. When you are looking at these charts, focus on how the data today compares to past market bubbles (i.e., how does current day data compare to similar data from periods like the late 1960s (pre-1970s recession); late 1990s (pre-Dotcom); and mid 2000s (pre-Great Recession).
Figure 9 highlights four different stock valuation gauges relative to history. As the yellow box indicates, all are at or near all-time extremes.
Figure 10, from Lohman Econometrics, show both earnings yield (stock earnings divided by stock price) and dividend yield (stock dividends divided by stock price) vs. performance of the S&P 500.
Figure 11 shows stocks returns against credit spreads. Credit spread is the interest rate on high yield (e.g., junk) bonds less interest rate on US Treasury bonds (e.g., high quality bonds). When spreads tighten – or go down – it can be an indication that stocks are overvalued.
Figure 12 indicates that stocks make up a record percentage of financial assets. When this percentage was high in the past, future stock returns were low.
Figure 13 indicates that stocks are a record percentage of GDP, or total economic output. Much like Figure 12 above, when the percentage was high in the past, future stock returns were low.
The final chart below warrants a little discussion. The price/earnings ratio (P/E ratio) is a common stock valuation. It is simply the price of a company’s stock over the earnings (e.g., profits) of the company. For example, if Alpha Corp has $2 per share of profits and its stock price is 30, it has a P/E of 15 [30 / 2]. P/E ratios, like other valuation metrics, allow us to compare stock valuations today vs. those in the past. Generally, the higher the ratio today vs. the past, the more overvalued the stock is today.
One knock on P/E is that earnings can be manipulated. There are various accounting “gimmicks” companies can play to increase or decrease their expenses, and therefore, increase or decrease their profits (earnings).
What is considerably harder for companies to manipulate is sales. Whereas companies can use various accounting tactics to massage expenses (and thus, earnings), revenue is revenue is revenue. A company either books a sale (revenue) or it doesn’t. Sales are much harder to massage. As such, the price/sales ratio (P/S ratio) is often used in concert with, or in lieu of, the P/E ratio.
The S&P 500 index consists of the 500 largest U.S. companies. Each company has its own P/S ratio. Envision a list with all 500 companies and their respective P/S ratios. Now, order the list from highest P/S to lowest P/S. Finally, make groups of 50 companies, starting with the companies that have the 50 highest P/S ratios, then the companies with the next 50 highest P/S ratios, etc. There are now ten groupings of 50 companies each. Figure 14 shows the P/S ratio of each one of these ten groupings. For example, the top, purple line represents the 50 companies with the highest P/S ratios. The light green line represents the 50 companies with the next highest P/S ratios.
The key takeaway here is that, over time, most companies have had low (e.g., < 2.0) P/S ratios. Currently, almost 400 companies, or 80% of the index, have P/S above 2.0. In other words, it’s not just a select group of stocks that is overvalued. It is pretty much the entire index.
A final thought of valuations. The rate of return of any asset is simply a function of three inputs:
The purchase price of the asset
The income produced from the asset
The timing of the income produced from the asset.
For example, assume you buy Alpha Corp today @ $100 (purchase price) and sell it one year (timing) from now @ $125 (income produced). Your return was 25% per year.
Add one layer of complexity to this example. Assume you buy Alpha Corp today @ $100, assume it pays a $5 dividend each year, and assume you sell it for $125 after two years. Now you cash flows look like:
Year 0 = -100
Year 1 = 5
Year 2 = 130 [5 +125]
The rate of return is the percentage that equalizes the outflows (e.g., -100 in Year 0) with the inflows. For brevity, let us skip the math and get to the result: 16.5% per year.
The math is not as important as the context: If we know the three inputs – purchase price, income, and timing, we can calculate the rate of return.
Based on S&P 500 dividend data compiled by multpl.com (https://www.multpl.com/s-p-500-dividend/table/by-year), the current annual dividend for the S&P 500 is $57.87. The dividend growth rate since 1900 has been 1.40% per year. Thus, we can project out a stream of dividends (input 1 = income) using the current dividend rate and the growth rate. For example, year 1 dividend = 57.87, year 2 = 58.68 [57.87 * 1.014], year 3 = 59.50 [57.87 * 1.014 * 1.104],…We know the timing of dividends is annual (input 2 = timing), and we know the current value of the S&0 500 is 4,384 as of July 12, 2021 (input 3 = purchase price). As such, we can calculate the expected average annual rate of return for the S&P 500 index:
In the base case of 1.40% dividend growth, the S&P 500 is expected to return just 2.72% annually forever.
If the dividend growth rate is adjusted as low as 1.00% or as high has 2.00%, the expected return for the index is 2.32% per year to 3.32% per year.
This analysis started off showing S&P 500 returns over various time periods. In no case was the average return 2.72%. Recall that the average S&P 500 total return since 1926 has been 10.41%. And yet, based on the current value of the S&P 500, the expected future return is only 2.72% per year.
One final thought on this analysis. Notice that interest rates have no bearing on the math. Again, the only inputs needed to calculate rate of returns are:
The purchase price of the asset
The income produced from the asset
The timing of the income produced from the asset
All the chatter that low interest rates justify high stock valuations is just that: chatter. But this chatter does have one very important role: It fosters a psychology of bullishness.
Sure, interest rates do not really matter when calculating the rate of return, but they do matter when investors have it pounded into their heads that low rates justify high valuations. Psychology is a hard factor to overcome, and it is the reason that the stock market continues to defy gravity. Empirically, stock valuations are high. Empirically, we should expect future returns to be below average. But momentum driven by the psychology of low rates can continue to drive the stock market higher.
Patience is a Virtue
It is hard not to chase returns. It is hard to maintain balanced portfolios when bonds have low interest rates and stocks seem to grow forever. But remember one thing: If stocks are overvalued today and they continue to climb higher, they will be even more overvalued tomorrow. The bigger they are, the harder they fall.
Maintain your discipline. Keep your portfolio balanced. We are here to help. We are only a phone call or email away.
Thank you for your continued friendship, partnership, and loyalty.
Respectfully yours,
Chris & Bob
-------------------------------------------------------------------------------------------------------------
The opinions voiced in this material are for general information only and are not intended to provide specific advice or recommendations for any individual. All performance referenced is historical and is no guarantee of future results. All indices are unmanaged and may not be invested into directly.
There is no guarantee that a diversified portfolio will enhance overall returns or outperform a non-diversified portfolio. Diversification does not protect against market risk.
Stock investing involves risk including loss of principal.
No strategy assures success or protects against loss.